>>which is better?
what matters is relevant results which address user/searcher intent. local search is segmented because local result sets are extremely different than traditional search results sets, enabling local consumers to interact with a display which inherently local business content and related functionality such as mapping, user-gen content, rating, distance scoring, product detail, et al. In this case, users are keep on the results pages longer than tradional search results, often interacting (sorting/filtering/calling/printing etc) and making decisions on the local serp without visiting a web site.
so if an se can serve more effective results through unique local result displays by segmenting them from traditional results, they would naturally attempt to integrate these results into traditional serps particularly for explicit local queries. this only serves to drive usage albeit slow to the pure local utilities.
we have long debated whether search segmentation (seperating the utility) is necessary, or whether a single search box can facilitate. the fact of the matter is 1)most local queries remain implicit therefor not enabling result alignment with intent (outside of geo-targ PPC) and 2) and most importantly, search result specialization will define search engines and se competetion for the coming years. put simply product shoppers, local search users, researchers,etc. have different needs when searching and interacting with content. today, prominant search tabs and prominent integrated results in tradition serps from segmented utilities are the best way to provide answers to users when intent/explicit needs can be acertained. therefore what we see today is a sound and stable integration of segmented display given that the overwhelming amount of search usage remains in the traditional search box. given the billions of local search being conducted and the need to promote usership in pure ls, this is the only logical answer.
where things will start to get interesting is when the se's have to face increased pressure from specialized vertical engines who are able to more effectively match intent and need with results and content. although arguably better for the user, these specialized se will face critical user mass issues leaving them, like many other search utilities without sufficient advertiser and user bases. how the major se's react to this will be a telling sign of the future of directional media. imo we havent even begun to see the major movement towards specialized results from se's. Geography was a logical early move. Industry will be next.